Depends on what you're trying to gauge - efficiency of a player or comparing styles of play.
Teams that want to play fast - like IU, UNC and UK - will have stats that likely will seem skewered from teams like Michigan St, Duke, etc . . . simply because a given style presents more opportunities. But compared side-by-side, they can indicate how well a program is able to execute a system against the results of teams playing a similar style.
We agree, then that rebounds per game isn't a useful metric regarding efficiency of a player. But I disagree that rebounds per game is useful comparing teams that want to play a similar style, and I also disagree with the premise that IU and UNC are comparable in terms of style.
With regard to comparing teams who play similar styles, there are still WAY too many factors that cause noise in the data. Most notably, defensive efficiency and shooting percentage, among others. So, even if I agreed that UNC and IU want to play the same style, the fact that IU was a good shooting team and only okay, overall, on defense, will depress their team rebounding numbers because of lack of opportunity. UNC, on the other hand, was a good defensive team with okay shooting, causing more opportunity, and noise, in the data.
All I've ever asked is WHY you wouldn't use pace-neutral stats...you've finally answered me, and even without personal insults. Thanks. I think there's more noise in the stats you used regarding rebounds and turnovers, and they are therefore less reliable than pace-neutral stats. I was interested to find out what your analysis would be using pace-neutral stats in those categories, but sadly, I think the thread got deleted anyway.
In terms of "playing fast," here are UNC's pace rankings the last 5 seasons: 73, 15, 19, 15, 10. An average of 26. IU's are: 151, 72, 63, 109, 108. An average of 101.
The difference, in terms of possessions per game, is 2, 3, 2, 3, 5, for an average of 3 possessions per game, in case you wonder if it makes a difference. There is some rounding going on there...didn't feel like using decimal places in my head. Again, an example of the noise that your per game stats can't account for--but not the only example.
To be sure, both teams play faster than average. But the evidence doesn't suggest that IU plays like UNC, or that they "want" to play like them. There are many teams, mostly small schools in lesser conferences, that want to play like UNC. Some even play faster. IU isn't one of them.
It's been awhile since I looked at them now, but I believe generally pace-neutral stats support the notion that 1) Crean is a championship caliber OFFENSIVE coach. 2) Crean's teams rebound well. 3) Crean is capable of coaching championship caliber defense, but has not done so consistently, or even half the time. 4) Crean's teams turn the ball over more than they should.
In terms of what I am counting as championship caliber, I call top 40 in Kenpom efficiency championship caliber, because no team has won the national championship in the one and done era, without being top 40 in BOTH categories, I believe. In fact, the winner is usually top 20 in both, but UConn's outlier national championship a couple years ago pushed the outer limit of what kind of team can be considered a national championship caliber team.
If these four conclusions, which are decidedly mixed, make me a "Crean apologist" or "settler" or whatever else, so be it. I'm not exactly giving out a ringing endorsement for a top tier coach, and I've never given him a ringing endorsement. I just prefer the discussion be honest and fair, rather than using noisy stats that might suit an agenda. You might recall, I didn't even know what the results would be when I first brought the issue up, and had to repeatedly insist that they might even help your argument.
And, it's been awhile since I looked at some of this stuff, so if my memory is off on any of it, feel free to correct me.